Collection method for knowing features of referenced resources

[azaroth42]

Desire expressed at the 2025 conference to be able to tell what sorts of resources would need to be rendered when processing Manifests.

e.g. if the client can't deal with 3d, then grabbing manifests that only have scenes would be pointless.

[azaroth42]

Related to #754

[azaroth42]

Rob as Rob: propose close, wontfix based on the same outcome for #754. There are too many dimensions to have to consider to either propose levels (groupings of features) or individual features for Presentation API resources.